Should you be transforming or weighting item responses in your evaluation surveys?

Abstract Information: Should an instrument be scored by simply adding the scores to the individual items in that instrument (unit scoring), or should items that correlate more strongly with the construct(s) measured by the instrument be weighted more heavily in the total score (weighted scoring)? Two recent papers reignited this decades-old debate. A similarly timeless controversy is whether scores should be transformed to induce normality prior to analysis, or whether the raw scores can be analyzed as is. Both of these issues were examined empirically, by scoring a popular instrument for assessing well-being with 2 different transformation methods and 17 different weighted scoring approaches – including random weighting. The well-being instrument is a strong candidate for weighted scoring and is known to produce scores that are not normally distributed. Attend this session to discover what was learned when the various scoring approaches were applied to two publicly available data sets. Can you rest easy, knowing that your survey responses have been scored appropriately, or do you need to change your quantitative methods?